The plasmid prophage of bacteriophage P1 can have as few as two copies per dividing cell. These are distributed to daughter cells by a precise mechanism, analogous to mitosis. We have been studying this system as a model for chromosome replication and segregation for the past twenty years. The P1 partition mechanism consists of a 2.1-kb region with and operon encoding parA and parB genes and a down-stream cis-acting site, parS. P1 ParA and ParB are the most studied members of a family of protein pairs encoded by a diverse group of plasmids and host chromosomes. They are responsible for chromosome segregation in several bacterial species. ParB binds specifically to parS, forming a protein-DNA complex. ParA is an ATPase that interacts with the core complex during partition. This year, we have made considerable progress in understanding the process by following the position and movement of the plasmid DNA during the cell cycle by microscopy in living cells. This is made possible by the properties of a GFP-ParB hybrid protein which loads onto the DNA as mutiple copies at the parS site, and produces a bright fluorescent focus within the cell. The results obtained are surprising. Two or more copies of the plasmid gather as a single focus at the cell center where they attached to some structure: probably the cell division apparatus. Partition consists of an explosive dispersal of these copies outward toward the cell poles that occurs just as the cell is dividing. As copies are always ejected in both directions cell division always results in two daughter cells, both of which contain at least one plasmid copy. Preliminary studies on mutant plasmids that are defective for partition suggest that they are unable to attach to the central structure, or in one example, that attachment occurs, but there is no explosive dispersal. Plasmid partition is an analogous process to chromosome segregation, and the existence of bacterial analogs of the ParA and ParB proteins suggests that there are mechanistic parallels between the two processes. Our continuing efforts to illuminate the process of chromosome segregation should therefore benefit greatly from our new findings concerning P1 plasmid partition. In addition, we have shown that the same fuorescence labeling of DNA sites described above can be adapted to following the dynamics of chromosome replication and segregation. The bacterial chromosome replicates from a unique origin and progresses bi-directionally, ending at a terminus region,approximately halfway around the circular chromosome. We have previously contributed to the finding that the replication forks are anchored to the cell center. Newly replicated DNA emerges from this central replication "factory" while the template chromosome is progressively drawn into it. Two very different models for how the sister chromosomes subsequently segregate have recently received support in the literature. In one, DNA replication drives segregation. The sequences destined for each sister chromosome are extruded away from the factory site in opposite directions, eventually forming two substantially separate masses in each cell half prior to cell division. The individual markers on the chromosome segregate away from their sisters progressively in the order in which they are replicated. In the second, the sister chromosome regions that emerge from the factory are paired, forming paired sister chromosomes that are analogous to sister chromatids in higher organisms. Late in the cell cycle, the sisters segregate away from each other as complete units by a process akin to mitosis. Using P1 parS sites integrated in the chromosome, and their cognate fluorescently labeled GFP-ParB binding protein, we have been able to follow the segregation of the origin and terminus sequences in living cells. The cell cycle was accurately monitored in the same culture using flow cytometry. We found that the origin sequence segregated into two separate sister copies soon after it was replicated. In contrast, the terminus sequences segregate much later, just before the cell divides. This observation provides strong support for progressive segregation during replication, and is not consistent with sister chromosome pairing models.